The present disclosure relates generally to the operation of autonomous machinery for performing various tasks at various industrial work sites, and more particularly to the volumetric estimation and dimensional estimation of a pile of material or other object, and the use of multiple sensors for the volumetric estimation and dimensional estimation of a pile of material or other object at such work sites. An application and a framework is disclosed for volumetric estimation and dimensional estimation of a pile of material or other object using at least one sensor, preferably a plurality of sensors, on an autonomous machine (e.g., robotic machines or autonomous vehicles) in various work-site environments applicable to various industries such as, construction, mining, manufacturing, warehousing, logistics, sorting, packaging, agriculture, etc.

A machine includes a frame, a plurality of traction devices supporting the frame, an engine and an operator cab mounted to the frame, an implement system configured to move the work tool to a desired position, position sensors, a tilt-rotate system to move the work tool to a desired orientation, orientation sensors, an operator interface, and a control module. The control module is configured to receive a model of the work tool, receive boundary inputs defining a virtual boundary, receive signals from the position sensors and the orientation sensors, receive implement control inputs from the operator interface, determine a position and orientation of the work tool based on the signals and the model, determine whether the work tool is approaching the virtual boundary based on the position and orientation, the boundary inputs, and the implement control inputs, and automatically prevent the work tool from crossing the virtual boundary.

An adaptive joystick preferably includes a rotatable cylinder bar, an outer base ring, an inner ring and an industrial joystick base. An adaptive controller receives an output from the adaptive joystick and outputs a control signal to a valve solenoid to control a hydraulic cylinder. Angle, depth and pressure sensors are preferably used to monitor a position of the hydraulic cylinder. The sensor outputs are fed into the adaptive controller. An inward wrist curl of the rotatable cylinder bar combined with a forearm pull rearward of the outer base ring are used to cause a digging motion. An outward wrist curl of the rotatable cylinder bar combined with a forearm push forward of the outer base ring are used to cause a dumping motion. A hand movement to the left is associated with swinging the excavator left. A hand movement to the right is associated to swinging the excavator right.

A distance calculation unit calculates a first distance that is a distance between a first bucket point being a point on a bucket and a target design surface representing a target shape of an excavation target. The distance calculation unit calculates a second distance that is a distance between the target design surface and a second bucket point. The second bucket point is on the bucket on a straight line passing through the first bucket point and is parallel to an edge of the bucket. A tilt control unit compares the first distance and the second distance to calculate a tilt control amount to rotate the bucket around a tilt axis.

Embodiments of a learning-based excavation planning method are disclosed for excavating rigid objects in clutter, which is challenging due to high variance of geometric and physical properties of objects, and large resistive force during the excavation. A convolutional neural network is utilized to predict a probability of excavation success. Embodiments of a sampling-based optimization method are disclosed for planning high-quality excavation trajectories by leveraging the learned prediction model. To reduce simulation-to-real gap for excavation learning, voxel-based representations of an excavation scene are used. Excavation experiments were performed in both simulation and real world to evaluate the learning-based excavation planners. Experimental results show that embodiments of the disclosed method may plan high-quality excavations for rigid objects in clutter and outperform baseline methods by large margins.

The disclosure relates to a method for monitoring and/or performing a movement of an item of machinery wherein the item of machinery comprises a movement device with a tool for picking up material, which comprises at least two components, each of which is movable via at least one actuator, and a control system by means of which the actuators of the movement device can be actuated by way of open-loop and/or closed-loop control. The method according to the disclosure comprises (i) detecting status information of at least two components, (ii) calculating torques that are applied to components, (iii) detecting torques actually applied to components, (iv) comparing the calculated and detected torques and determining a force vector actually applied, and (v) executing an action depending on the calculated force vector. The disclosure also relates to an item of machinery and a computer program product for executing the method.

An excavator includes a lower traveling body; an upper turning body turnably mounted to the lower traveling body; an attachment that is attached to the upper turning body; and a power engine that is mounted to the upper turning body. The excavator is configured such that, before a low-load operation by the attachment is started, a rotation speed of the power engine is reduced.

A system and method are provided for operating a work machine comprising a laser receiver and an implement for working a terrain. Responsive to movement of the laser receiver, a laser reference is received at a plurality of positions relative to a transmitting laser source, wherein the laser reference corresponds in slope and direction at a defined elevation offset with respect to a target surface profile of the terrain being worked. A plane of the laser reference is determined from data points corresponding to the plurality of positions at which the laser reference is received, and movement of at least the implement is controlled with respect to at least the determined plane of the laser reference and the defined elevation offset.

A point generating unit of a mobile terminal generates teaching point information associating a teaching position that teaches the work machine a position of the attachment in a series of movements to be performed by the work machine with orientation information indicating a target orientation at the teaching position, based on a slewing angle of an upper slewing body and on orientation information on the attachment. A point changing unit changes the generated teaching point information. When the teaching point information is changed, an instruction generating unit of the work machine generates an automatic operation instruction for automatically operating a slewing device and the attachment, based on the changed teaching point information. An operation control unit of the work machine automatically operates the slewing device and the attachment, based on the automatic operation instruction.

A shovel includes a plurality of driven elements, a plurality of actuators configured to drive the plurality of driven elements, and a hardware processor configured to, in response to detecting that two or more actuators of the plurality of actuators are synchronously moved, prohibit a motion of another actuator of the plurality of actuators that is different from the two or more actuators of the plurality of actuators.